Asymptotically Stabilized Sliding Mode Output Feedback Control of n-DOF Robot Based on ESO With Neural Network

This article focuses on developing a nonlinear motion controller for n-DOF robots to achieve asymptotically stable tracking performance despite the presence of external disturbances and parameter uncertainties, a novel controller named asymptotically stable sliding mode output feedback control based on expanded state observer (ESO) and neural network is proposed. A novel nonlinear state observer based on ESO and neural network only driven by observation error of position in the system is designed. The composite observer is used to observe the joint velocity and uncertainties, including the uncertainties of system parameters and external disturbances. Based on the observation of uncertainties and velocity, the dynamic switch function, and feedforward compensation is carried out in the controller, an asymptotically stable sliding mode output feedback control with the integral term of the symbolic function of sliding mode variable is designed. Finally, the Lyapunov stability analysis method is used to prove that the designed observer can achieve bounded stable observing performance, and the controller can achieve asymptotically stable tracking performance, which is very important for high-precision motion control of robot systems or any other nonlinear systems. A large number of comparative simulations and experiments verify the high-precision tracking performance of the proposed control strategy.